Intel launches high-performance AI chips designed to compete directly with AMD in the AI accelerator market.

Intel has been facing significant challenges in recent years, particularly following the launch of the iPhone with Apple silicon and the rise of an AI and data center chip industry largely dominated by Nvidia and AMD. Rather than ceding ground, the company has announced the release of its Xeon 6 chips and new Gaudi 3 AI accelerators designed to address these competitive pressures.

The Intel Xeon 6900 P-core series offers up to 128 cores for extremely intense AI workloads, doubling the performance of its predecessors while providing higher core counts, more memory bandwidth, and embedded AI acceleration. The new Gaudi 3 AI Accelerator chips, aimed at generative AI applications, come equipped with 64 Tensor CPU cores, eight matrix multiplication engines, and 128GB of HBM2e memory, offering up to 20% more throughput and twice the price and performance over Nvidia's H100 chips for LLaMa 2 70B inference.

Beyond designing and manufacturing the chips, Intel has also partnered with Dell and Supermicro to develop "co-engineered systems" specifically tailored to the AI infrastructure needs of those companies. Dell is co-engineering RAG-based solutions that leverage both the Gaudi 3 and Xeon 6 technologies.

Meanwhile, AMD has emerged as the biggest beneficiary of Intel's struggles in the AI chip space, designing powerful processors manufactured by TSMC. AMD plans to release its fifth-generation EPYC Turin 3nm data center chips in the second half of 2024, based on its Zen 5 architecture and featuring up to 192 cores and 384 threads, positioning it to compete directly with Intel's latest offerings.

In the closely-watched Top500 list, the Intel-powered Aurora supercomputer, which is not yet fully operational, came second behind the AMD-powered Frontier but took first place in an AI benchmark, suggesting Intel's innovations may yield real-world performance gains despite the intense competitive landscape.